A Gap in OSCE Conflict Prevention? : Local Media and Inter-Ethnic Conflict in the Former Soviet Union

No. 645 – 2003

Norsk

Utenrikspolitisk Institutt Norwegian Institute

of International Affairs

Indra Øverland

[ 645 ^] Paper

A Gap in OSCE Conflict Prevention?

Local Media and Inter-Ethnic Conflict in

the Former Soviet Union

Besøksadresse:

Addresse:

Internett:

E-post:

Fax:

Tel:

Alle synspunkter står for forfatternes regning. De må ikke tolkes som uttrykk for oppfatninger som kan tillegges Norsk Utenrikspolitisk Institutt. Artiklene kan ikke reproduseres - helt eller delvis - ved trykking, fotokopiering eller på annen måte uten tillatelse fra forfatterne.

Any views expressed in this publication are those of the author. They should not be interpreted as reflecting the views of the Norwegian Institute of International Affairs. The text may not be printed in part or in full without the permission of the author.

Grønlandsleiret 25 Postboks 8159 Dep.

0033 Oslo www.nupi.no [email protected] [+ 47] 22 17 70 15 [+ 47] 22 05 65 00

Indra Øverland

Prevention?

Local Media and Inter-Ethnic Conflict in the Former Soviet Union

[Abstract]

Contents... 1

Introduction ... 2

Conflict prevention... 3

The media in post-Soviet conflicts ... 5

The OSCE ... 15

The media in OSCE annual reports ... 17

The HCNM and the RFOM... 19

Conclusions: closing the gap between conflict prevention and the media ... 25

Appendix 1. Media outlets in the former Soviet Union ... 29

Appendix 2. Use of the word ‘media’ in OSCE annual reports ... 30

Appendix 3. HCNM and RFOM percentages of OSCE budgets ... 31

Bibliography... 33

The objectives of this paper are twofold. Firstly, the article aims to show that local media have been of great importance in the escalation of inter-ethnic conflicts in the former Soviet Union. Secondly, it aims to uncover whether the OSCE (Organisation for Security and Co-operation in Europe) has given the media a corresponding role in the efforts at preventing the escalation of those conflicts.

The starting point for the article is the hypothesis that the OSCE initially neither had the capacity nor fully saw the necessity of including local media at the forefront of its conflict prevention agenda, and that this debilitated the organisation as an agent of conflict prevention. However, towards the end of the article I also argue that during the late 1990s there have been indications that this media gap in OSCE conflict prevention may be filled. During the past few years media issues have increasingly come to preoccupy the OSCE, chiefly in connection with human rights issues and freedom of speech, but to some extent also as an element of conflict prevention.

This argument involves, rather than focusing on a few case studies of the role of media in specific conflicts, a broad sweep across what used to be the Soviet Union and is currently a motley group of contiguous states. The advantage of this broad sweep is that it makes it possible to capture the full range of post-Soviet developments, including both the heterogeneity and homogeneity of the region. In addition, it is interesting to look at the region as whole as a relatively self-contained chunk of the OSCE’s operative area.

Inevitably, this also means that the depth of the study in relation to each case is limited. This is not a matter of choice. Covering the entire region in detail and at first hand would have demanded an entirely different scale of resources than have been available to this project. On the other hand, I have striven for methodological creativity in the use of sources that are partly sec- ondary and diverse. The main sources have been studies of individual post- Soviet conflicts, Soviet census statistics, OSCE budget numbers and official OSCE documents. Some of the OSCE documents have been subjected to word frequency analysis, in which the frequency of a word is used to deter- mine the importance of the topic with which it is associated for the organisa- tion. In addition I draw on altogether six trips to three different countries in the region: Russia (1997, 1998, 1999, 2000), Azerbaijan (1998) and Ukraine (1999); altogether about six months of travel.

By drawing on these diverse sources of information, many of which leave much to be desired but still provide at least circumstantial evidence, it is my hope that a more general argument can still be made about the OSCE, con- flict prevention and the media throughout the region.

Following Björkdal (1999: 57) I define ‘conflict prevention’ as ‘actions and institutions that are used to keep political disputes between and wthin states from becoming violent, and to hinder existing violent conflicts from escalat- ing’ (cf. Stedman 1995: 14). Defined in this broad manner, conflict preven- tion can be applied both to (1) the initial outbreak of violence, (2) its escala- tion or (3) the relapse of old conflicts into it. Accordingly, it has the broadest meaning of the ‘contraceptive family of terms’. Among these terms are ‘pre- ventive action’, ‘preventive engagement’, ‘preventive deployment’, ‘conflict prevention’, ‘crisis prevention’ and ‘preventive diplomacy’ (Saunders 1998:

63). ‘Preventive diplomacy’ is the original and most important family mem- ber.

‘Preventive diplomacy’ stems from UN Secretary-General Dag Hammarskjöld, who thought of it as the UN insulating conflicts from the great power rivalry of the Cold War (Steiner 1998: 19; James 1999: 43, 44;

Jentleson 2000b: 9–10). The concept rebounded back to prominence with the publication of Boutros Boutros-Gali’s Agenda for Peace in 1992, an entire section of which is dedicated to preventive diplomacy. Boutros-Gali identi- fies four components of preventive diplomacy: fact-finding, confidence- building, early warning and preventive deployment (Boutros-Gali 1992: 11–

19; cf. Väyrynen 1996: 22).

I have chosen to use the term ‘conflict prevention’ rather than ‘preventive diplomacy’ because it reflects the broad concept I am working with more faithfully, and the word ‘diplomacy’ can easily be understood more narrowly than necessarily (cf. Ginifer and Eide 1999a: 7). Even Boutros Boutros-Gali has later realised this limitation, causing him to favour ‘preventive action’

over ‘preventive diplomacy’ (Sokalski 1999: 74).In practice, however, pre- ventive diplomacy and conflict prevention overlap to such a great extent that they can be, and often are, understood synonymously (e.g. Bercovitch 1998:

128).

Conflict prevention became particularly popular in the wake of the Agenda for Peace, in particular in the USA. Among other things, the New York-based Carnegie Corporation set up a Commission on Preventing Deadly Conflict, the US Institute of Peace established a study group on pre- ventive diplomacy, the Carnegie Endowment for International Peace pro- posed the creation of a global crisis team, and the US Council on Foreign Relations maintained a Centre for Preventive Action to study and test con- flict prevention (Stedman 1995: 14). In the vein of the burgeoning literature on conflict prevention published during the 1990s, Lee Hamilton (2000: xi) wrote that ‘no issue facing the world today deserves our attention more than conflict prevention. We need to foster a sense of urgency, a new way of thinking that gives precedence to the prevention and not simply the manage- ment of conflict, to avoid disaster rather than merely dealing with its cones- quences.’

Among the most central of the many works on conflict prevention from the 1990s is Michael Lund’s (1996) Preventing Violent Conflicts. The policy of conflict prevention in general and Lund’s work in particular is highly praised in some quarters (e.g. Bercovitch 1998: 128), and scolded in others (e.g. Stedman 1995: 15). In a review of works on security issues and conflict prevention, Mark Clark (2000: 158) states that ‘the realism of the Cold War era has evaporated, only to be replaced by a myopic idealism that descends like a fog’ (see Charles-Philippe David 1999 for a more balanced critique of conflict prevention).

One criticism of the concept is that it is a simplistic resort to an appealing medical metaphor that has little to do with the reality of political conflict.

Such metaphors certainly do often crop up in the literature on conflict pre- vention. Jentleson (2000b: 3) provides a classic example: ‘It is the same logic as preventive medicine: don’t wait until the cancer has spread or the arteries are fully clogged’ (for other examples of medical metaphors see Bercovitch 1998: 129; James 1999: 45; Brown and Rosencrance 1999:

cover). Relying on argument or analysis by analogy is generally question- able, but not enough to undermine the general argument for conflict preven- tion.

More specific critiques of conflict prevention question the ability of would-be conflict preventors both to single out the worst conflicts in advance (the problem of the ‘signal-to-noise’ ratio), and to muster the politi- cal and financial resources to do anything about them in time (which depends upon the CNN-curve, i.e. public international attention) (Jentleson 2000b: 11, 13; cf. Tharoor 1999: 15, 34): ‘There is little basis for optimism in the ability of social science to precisely forecast the outbreak of violent domestic conflicts’ (Stedman 1995: 16).

Related to these problems is that of the long-term public recognition of conflict prevention as a meaningful activity. Part of the problem is that it must often be carried out discretely in order not to exacerbate the conflicts it aims to assuage. In addition, its success is inextricably tied to the absence of precisely those large-scale conflicts that might draw the necessary public attention to generate the required political pressure to spur politicians and diplomats into serious conflict prevention. As has been written of the OSCE High Commissioner on National Minorities: ‘What, it may be asked, has he achieved? The trouble is, his performance has to be judged largely by what has not happened’ (The Economist 1999: 36). Or as Lund (1996: 164) put it:

‘How does one prove that one prevented something that didn’t happen?’

Underlying this practical problem in conflict prevention is the philosophical problem of determinism and counterfactuals in the study of history.

However, these deeper problems relate to conflict prevention as a total concept. Rather than attempting to assault or salvage the concept as a whole, I start from the assumption that conflict prevention is possible and meaning- ful, and go inside strategies of conflict prevention to show that media-related measures are integral to any successful attempt at conflict prevention in modern societies.

It is well-known and often mentioned in the West that the media were at the centre of the Velvet Revolutions in the former East Bloc. Images of peaceful but resolute crowds engulfing national television stations are familiar and often conjured up in both academic and journalistic writing on the former East Bloc (e.g. Hurlburt 2000: 94). The Velvet Revolutions are implicitly envisaged as struggles between entrenched political elites and frustrated peoples, over information, truth and world view. Although the frustrated peoples have mostly won those struggles, that is not to say that the elites did not understand the importance of the media. The media were in fact also ‘the vehicle through which the Soviet elite attempted to garner support for the policy of perestroika, holding out glasnost as the key to a brighter future’, in the first place (McCormack 1999a: 7).

Here, however, many Western observers’ understanding of the media in the former Soviet Bloc tends to stop. They tend to envisage former Soviet media as the mouthpieces of homogeneous and unitary societies controlled by or struggling against their leaders and elites, and fail to take in the cul- tural, linguistic and ethno-political diversity of post-Soviet societies that is inevitably reflected in the media. It is tempting to see in this the perpetuation of the Western layman’s tendency to refer to the entire Soviet population as

‘Russian’, and of the Sovietologist’s blindness to the centrifugal force of economic slump combined with ethnic discontent in the 1980s.

The first prerequisite to understanding the post-Soviet media landscapes is to realise the sheer number of media outlets. To my knowledge a full and systematic statistical overview of media in the former Soviet republics does not exist. The main reason for this is the instability of the post-Soviet media landscapes, rendering any estimate unreliable. It is nonetheless possible to gather an impression from the rudimentary numbers that do exist. Combined, McCormack’s Media in the CIS and Schmid’s Understanding the Media in the Baltic Countries provide a rough but useful overview. Almost all the numbers in the two works are qualified with ‘approximately’, ‘about’ or

‘roughly’, and statements that such numbers of officially registered media greatly exaggerate the numbers of media actually operating on a day-to-day basis. Nonetheless, the numbers still convey the richness of the post-Soviet media flora.

Russia and Kazakhstan, both among the largest post-Soviet states, are registered as having respectively around 12,900 and 2,100 media outlets each in 1998, including news agencies, magazines, newspapers and televi- sion and radio stations (McCormack 1999a: 137, 219). Even the smallest of the states mostly have numerous outlets, for example Moldova with 357 and Estonia with 572 (Unt 1998: 43; Dodd 1998a: 56). The main exceptions are Belarus and Turkmenistan, which have the most oppressive censorships and the bleakest media landscapes (McCormack 1999a: 276; OSCE 2000a: 61);

and small separatist regions that are not (yet) fully-fledged states, such as Abkhazia, South Ossetia and Nagorno-Karabakh (McCormack 1999a: 6).

The media landscapes in the former Soviet republics are further compli- cated by the activities of foreign media offering both print and programming in English, Russian and the vernaculars. Voice of America and Radio Lib- erty are particularly well-known for their activity in the region, but there are numerous others as well. The situation in Tajikistan, an ethnic patchwork on Afghanistan’s unstable northern flank, is illustrative: the BBC, Voice of America, Radio Liberty and the Voice of Free Tajikistan (funded by Iran and broadcast from the Kunduz Province in northern Afghanistan) all produce radio in Tajik (McCormack 1999a: 258). In addition, the first three are also available in English, and a number of Russian television and radio channels also broadcast to Tajikistan. Finally, the population of Tajikistan can also pick up several channels from neighbouring Uzbekistan, Kyrgyzstan, China and Pakistan.

The numerous media in Russia proper are also features of the media land- scapes of the newly independent states, partly explaining the lower numbers of domestic media per capita in some of the newly independent states. In a few cases the authorities of the newly independent states have however taken control over Russian media input, as in Turkmenistan, where the main (and Kremlin-friendly) Russian channel ORT is recorded and censored, before being retransmitted to the general populace by the national broadcaster TMT when it suits the Turkmen authorities. In 1998, for example, the Turkmen authorities were displeased with ORT and cut transmissions from 18 hours to five (McCormack 1999a: 276). For a full overview of estimates of the num- bers and kinds of media in the former Soviet Union, see Appendix 1.

Against the claim that the media are of great importance in post-Soviet societies it might be contended that many of the media outlets are in fact quasi-newspapers with limited print runs, infrequent editions and narrow readerships; and that at least most post-Soviet print media have either col- lapsed entirely or at least diminished greatly in societal importance as a result of lacking financing from investors and diminished purchasing power among readers.

These problems certainly do apply to the post-Soviet media landscapes.

The development of the print media in Ukraine during the first half of the 1990s is a case in point. From September 1990 to May 1992, concurrent with the Soviet collapse, there was a frenzy of new media outlets. During those two years, 2,869 print media were registered, many with ambitious print runs. Subsequently, between 1992 and 1995 their combined circulation halved every year, dropping precipitously from 63.5 to 8.44 million copies (McCormack 1999a: 301).

Also the Latvian case illustrates the limitations of many post-Soviet pub- lications. In 1996, 207 out of 443 titles were ‘small, rapid-growth news- papers published by individual factories, firms and colleges. These … [were]

mostly published irregularly, or one to three times a month, and have four our sometimes only two pages’ (Schmid 1998: 169).

Inevitably, the economic pressure on the print media has strengthened the electronic media, also rendering them more important as instruments for

‘promoting the aims of the conflicting parties’ (Reljic 1999: 16). However, in spite of these weaknesses and imbalances in the media landscapes, it would be entirely wrong to portrait the former Soviet Union as demediafied.

All kinds of media have retained societal potency. Although obviously the print media are much less read than previously, as indicated by the Ukrainian

example, they remain important for the political dynamics and the formation of public opinion. About 18 per cent of the population in Ukraine still read newspapers regularly, and only 14 per cent do not read them at all (McCormack 1999a: 306). Several of the newspapers remain among the most important sources of information for the population. For example, the Russian-language weekly Zerkalo nedeli is read and taken seriously by most parts of the elite in Ukraine. With a circulation of only about 30,000 copies it is not even the largest newspaper in the country (cf. Silski visti, which enjoys a circulation of 522,000) (McCormack 1999a: 305, 308).

Considering the relatively numerous media and vigorous media land- scapes in the post-Soviet states, Velvet Revolution images of thronging crowds reasserting popular control over the television station can easily be misleading. There are many different television stations and other media in each of the states. It is unlikely that they all represent the unified voice of the entire population in each state.

That is my argument about the relatively deep media penetration of post- Soviet societies, multiplicity of post-Soviet media outlets and heterogeneity of post-Soviet macro media landscapes. That argument must, however, be combined with a second and contrary argument about the narrowness of many ethno-politically defined media micro landscapes in the region. In order to understand the role of the post-Soviet media in inter-ethnic polarisa- tion, it is necessary to move beyond sheer numbers of media outlets, to the factors that divide those media between different groups and political dis- courses. The most important dividing factor is inevitably language.

Comprehensive statistics of the entire Soviet population are available from the decennial Soviet censuses. Each census details among other things mother tongue and ethnic identity. The linguistic reliability of the census numbers on mother tongue is often questioned, since respondents frequently would claim a mother tongue that matched their ethnic identity, while in fact they might be far more proficient in another language, often Russian (Tishkov 1997: 87–88). For the purposes of my argument this statistical flaw is, however, not fatal. The same ethnic identity could drive people to adhere increasingly to their ‘mother tongue’ media, and is probably at least as powerful as the countervailing force of actual language proficiency. For example, even those Azeri who in fact initially spoke better Russian might adhere to Azeri media in a situation of conflict and nationalist fervour.

The last Soviet census was carried out in 1989. It is striking that of all the autonomous republics in the Soviet Union, the Checheno-Ingush Autonom- ous Republic had the highest percentage of respondents who considered the language of their titular nationality their first and most important language (Lapidus 2000: 43; Tishkov 1997: 96). This in itself does of course not explain the conflict in Chechnya. Several other groups in the Soviet Union had numbers close to those of the Chechens, and still have not experienced major unrest. However, the numbers from the census are still interesting pointers, and high percentages of most of the groups that have been involved in inter-ethnic violence in the 1990s counted their own ethnic group’s lan- guage as their mother tongue.

Moldova, with a population of only 4.4 million and the approximate size of Belgium, provides a good illustration of how even the smallest of the post-Soviet states are sufficiently linguistically diverse to provide for divided media landscapes. In 1998 there were 213 officially registered newspapers in

Moldova, of which 90 were published in Moldovan, 95 in Russian, 18 in both Moldovan and Russian, two in Gagauz (the language of a Turkic minor- ity), two in Bulgarian and six in other languages (McCormack 1999a: 185).

The most important feature of the Moldovan media landscape is of course the divide between Russian and Moldovan media outlets, which underpins the division of Moldovan society into two opposing halves.

The Moldovan case illustrates well how the adherence of cohabiting eth- nic groups to different languages in a conflictual situation can provide the basis for increasingly divergent interpretations of that situation (Tishkov 1997: 99). Once groups are sufficiently suspicious of each other, they can ignore the media of the other group, and rely increasingly on those using the language of their own group. In this way the basis is laid for the develop- ment of divergent ethno-political discourses with opposing truths, histories, guilt and unavenged crimes.

Even Lithuania, one of the most mono-ethnic of the post-Soviet states, has the basis for a linguistically divided media landscape. The population is 80 per cent Lithuanian, nine per cent Russian and seven per cent Polish.

Nonetheless, 48 print media are published in Russian and nine in Polish, enough to support diverging ethno-political discourses (Schmid 1998: 171).

The relation between the Abkhaz and the Georgian media landscapes pro- vides a good example of the mechanisms involved in the development of such divides when violence erupts. According to Chania (1999: 43) and Tushba (1999: 65), the state-controlled Abkhaz television and radio broad- casting company has a virtual monopoly on providing the Abkhaz public with information, and neither other Caucasian nor Russian media penetrate the breakaway republic, with the exception of Russian television. Chania and Tushba represent the Abkhaz perspective and emphasise the informational isolation of Abkhazia. In contrast, Pachkoria, who represents the Georgian perspective, claims that Georgian radio is in fact audible in Abkhazia, and that Georgian television can be watched in at least the Gali district of Abkhazia (cited in McCormack 1999e: 80). He does not, however, argue that the Abkhaz actually watch Georgian television or listen to Georgian radio.

Instead it is implicit in his account that the Abkhaz do not take in Georgian media because they do not want to do so.

For the purposes of the argument in this article both points are equally helpful. Chania and Tushba’s point shows that a media divide and infor- mational isolation indeed are aspects of the Georgian–Abkhaz conflict. Con- currently, Pachkoria’s point indicates how such a divide is not solely due to circumstance, differing languages and externally enforced isolation, but can also be strengthened by voluntary self-isolation and disregard for informa- tion from the other side of the ethnic divide.

Fawn (2000: 47) describes how the almost total divide between Georgian and Abkhaz infostreams results in the construction of distinct realities by the two ethnic groups:

The violent Abkhaz-Georgian stalemate rests on the opposing interpretations of the war of 1992–93 … the primary obstacles to negotiations derive directly from the differing perceptions of the origins and course of the war. The Georgians insist that their co-ethnics who were displaced from Abkhazia in autumn 1993 and who have lived as refugees thereafter must be allowed to return in full. The Abkhaz, by contrast, assert that a political settlement is a precursor to the refugee

question; therefore, the status of the de facto Abkhaz state needs at least to be acknowledged. Furthermore, the Abkhaz authorities insist that they be allowed to vet the returnees for ‘war crimes’ they have committed against the non-Georgian population of Abkhazia.

A similar informational divide feeds the Georgian and Ossetian perspectives on South Ossetia:

Even worse things happen when we are instigating and fuelling inter-ethnic con- flicts as we were doing until quite recently … we were shaping an image of the enemy. Journalists do these things to please politicians and because it is easy to write and present facts in a controversial way with clear-cut good guys and bad guys. It happens because we are stewing in our own juices, because we are sepa- rated from each other (Gingikhashvili 1999: 74).

Of course, distinct language groups in themselves are not enough for the division and antagonisation of a media landscape. Even in those cases where a multilingual media landscape is combined with already tense relations between groups, it cannot be said that the media always play an aggravating role. Rather than producing predetermined reactions to social processes, media professionals make choices entailing an active, though not singularly determinant, role for the media. It has been said ‘that journalists are the mediators between the government and the public. More than that, they influence the development of society. This is absolutely true’ (Kharebov 1999: 60). Concerning the conflict over South Ossetia, Kharebov (1999: 61) argues that the conflict between Ossetians and Georgians developed through three stages: journalistic, political and military – and moreover that the jour- nalistic phase incepted the entire conflict.

Representing the Karabakh-Armenian perspective on the Nagorno- Karabakh conflict, Bagdassarian (1999: 75) takes this argument even further, positing journalist not only as the potential abettors of conflicts, but as their actual perpetrators:

Such conflicts always begin with words. The Russian poet Tyutchev said: “We cannot foresee what will be the response to our words.” This is very actual dur- ing times of conflicts, because a word carrying a negative charge is more danger- ous than a bullet. Yesterday my Azerbaijani colleague Arif Aliev and I discussed a very interesting topic: what would happen if journalists were in charge of resolving the Karabakh conflict? And we came to the joint conclusion that it would be much worse.

When the media do take on a role in ethnic conflicts, they have two main types of influence: explicit and implicit. Firstly, they can influence their viewers, listeners and readers through explicit ethnocentricity, blatantly con- demning other ethnic group(s) and directly inciting their own. McCormack (1999e: 88) writes about such explicit media influence:

Ethnic chauvinism is omnipresent in the media during times of nationalist ten- sion and conflict. It is a major instrument of paving the ground for mass scale violence. Merciless propaganda is a factor of sometimes decisive importance for the outcome of armed conflict. Recent wars in the former Soviet Union and the disintegrated Socialist Yugoslav Federation proved this.

Similarly, Lapidus (2000: 56) states about the role of the Russian media in the first war in Chechnya:

The combination of information and disinformation spewed out by official Rus- sian sources, as well by nationalist propagandists, tended to neglect the obvious collusion between Russian and Chechen elites that had contributed to the situ- ation and portrayed the Dudayev leadership as nothing more than a criminal con- spiracy without popular support or legitimacy, manipulating separatist political slogans to disguise its real goals. It was, moreover, permeated with ethnic stereo- typing and scapegoating that came close to treating Chechens as a criminal popu- lation (cf. Andrei Pontkovsky, cited in Herd 2000: 32).¹

There is no lack of examples of the direct negative influence of the media through explicit exhortations. Nonetheless, implicit perspectives, stereotypes and judgements are equally effective and even more widespread. For example, referring to the conflict between Georgia and Abkhazia, Chania (1999: 45) states that ‘both we and they thoroughly underline the national colours of any crime made by someone from the opposite side. Both we and they painstakingly conceal human rights violations if they were done by our own people’. By recounting events with an ethnic hue, the media can polar- ise groups.

Much implicit media influence is subtler. By projecting skewed images of history, geography and culture, the media can support differing realities.

Fawn (2000: 52) writes about how the Georgian authorities and media try to eliminate the Abkhaz issue by avoiding referring to it or anything that might be associated with Abkhaz demands:

While the loss of Abkhazia pains both the population and the Shevardnadze gov- ernment, both engage in cognitive dissonance. From the national currency to bus shelter advertisements, the map of Georgia is shown as including Abkhazia. New highway signs across Georgia give the distance to Sukhumi, even though a Georgian would never be able to drive into Abkhazia.

This ‘cognitive dissonance’ is in turn aided by the official use of geographi- cal pseudonyms. Georgian media refer to ‘Samochablo’ and ‘Shigo’, refus- ing to recognise ‘South Ossetia’ both as a name and an administrative unit, seemingly fearing that that would somehow entail recognising the Ossetian cause (Kochiev 1999: 78).² An important component of the Georgian–

Abkhaz conflict lies in the ensuing mutual misrecognition: ‘For us, these are negotiations between representatives of two republics, Abkhazia and Georgia. For them, these are negotiations of the Georgian leadership with the separatist regime … When covering the issue of terrorist groups operating in Gali district – we talk about terrorists, they talk about guerrillas’ (Chania 1999: 43).

1 On the media front, the first war in Chechnya (1994-96) was disastrous for the Russian side (cf. Lapidus 2000: 48). In the second war, in contrast, the Russian authorities were acutely aware of the importance of the media for the success of the campaign and man- aged to maintain a more unified supportive stance among the Russian media (Herd 2000:

31, 32).

2 Cf. the use of contrasting weather maps by Serb and Croat television, reflecting ethno- political stances. While Serb television tends to give the weather for the entire former Yugoslavia, Croat television tends to show only Croatia (Reljic 1999: 18).

When opposing groups do try to engage in dialogue, the established cog- nitive divergence can create further trouble, in terms of an inability to recog- nise and relate to each other. For example, in the case of the conflict between Armenians and Azeris over Nagorno-Karabakh, ‘even details such as the question of whether representatives from Nagorno-Karabakh should have a nameplate at the negotiating table, and what title might be inscribed on it, were loaded with symbolism and were thus very sensitive. Essentially, neither side was yet ready to find a compromise’ (Maresca 2000: 80). Simi- larly, Chechen nationalists insist on referring to their breakaway republic as

‘Ichkeria’ rather than ‘Chechnya’, and its capital as ‘Djohar’ rather than the Russian ‘Grozny’ (which means ‘terrible’, ‘formidable’ or ‘menacing’).

Hence the compromise formulation used in documents such as the 1997 agreement On Peace and the Principles of Relations between the Russian Federation and the Chechen Republic of Ichkeria (cited in Lapidus 2000: 50, 51).

The post-soviet media should of course not be stereotyped into only being producers of hate speech and the bloodthirsty ‘stenographers of those in power’, to borrow a phrase from Reljic (1999: 19). While local media have played problematic roles in virtually all the major post-Soviet conflicts, there are both exceptions and progress in some cases. For example, Armen- ian newspapers, radio and television have recently given Azeri politicians and diplomats the opportunity to express themselves without editing to the Armenian public (Verdyan 1999: 49). During the first war in Chechnya several Russian media outlets helped end the war by conveying its brutality and meaninglessness to the Russian public.

The contradiction is not absolute between a strong sense of ethnic belong- ing among journalists on the one hand and balanced journalism on the other, and the prevalent negative role of local media in post-Soviet ethnic conflicts not inevitable, in spite of ethno-linguistic differences.

I believe that journalists from the Caucasus always feel their ethnic identity and nationality. Nevertheless I would like very much that at the same time journalists would realise that they best protect the interests of their nation and fulfil their commitments as a citizen when they professionally carry out their journalistic commitments, i.e. when they provide maximally trustworthy and maximally reli- able information of the highest quality. But when they try to be politicians, then as journalists they are poorly fulfilling their obligations to their compatriots’

(Navarsardian cited in McCormack 1999d: 82)

An important aspect of any conflict prevention strategy involving the media will therefore be to disentangle senses of ethnic identity, professional jour- nalism and what is indeed ‘good’ for one’s people.

Before such a strategy is possible at all, it is necessary to understand the role of the media in ethnic conflict. At the most basic level, the understand- ing of the role of the media in post-Soviet conflicts depends on how ethnicity and ethnic conflict are understood at the theoretical level. Here the distinc- tion between primordialism and instrumentalism is of some relevance. As perspectives on ethnicity and nationalism, primordialism and instrumental- ism have important repercussions for how ethnic conflict is understood and dealt with. By drawing on these theories it will hopefully be possible to avoid what was once described by Motyl (1992, 255) as the ‘huge chasm’

between studies of ethnicity and nationalism in the former Soviet Union and theories of ethnicity and nationalism in general – a ‘chasm’ that has yet to be properly bridged.

The term ‘primordialism’ derives from Shils and Geertz (Grosby 1994:

166; Linz 1985: 204).³ Shils (1957: 142) referred to primordial ties as the social bonds within families and in close personal relationships, contrasting them to more impersonal relationships. Geertz (1973: 159–60) brought the term closer to its current usage in the study of ethnicity and nationalism by examining ‘the lifting of such ties to the level of political supremacy’. It has since taken on the general meaning of sympathising with ethnic and nationalist movements, lending them credibility and seeing them as the inevitable outcome of distinct cultures (Gellner 1997: 93; Eller and Coughlan 1993: 187; cf. Grosby 1995: 148).

Primordialism is generally posited as the opposite of various deconstruc- tive and critical approaches to ethnicity and nationalism, often referred to as

‘modernism’, ‘constructivism’, ‘contextualism’ or ‘circumstantialism’

(Calhoun 1997: 30; Kellas 1991: 56; Schmidtke 1998: 25; White 1999: 789).

Depending on which of the opposing approaches it is pitted against, the term

‘primordialism’ has various meanings relating respectively to psychology, history or politics. Thus ‘primordialism’ is sometimes taken as meaning the notion that ethnic or national identity is fundamental, natural or inevitable (e.g. Stack 1986: 1), or that political action on the basis of such identity is.

Yet other times by ‘primordialism’ is meant the ineffability of ethnic ties (Hechter 1986: 13; Smith 1994: 706; Eller and Coughlan 1993: 187) or simply ethnocentrism or xenophobia as such (Schmidtke 1998: 30).

However, the approach most often posited as the opposite of primordial- ism is instrumentalism (Stack 1986: 5; Rex 1996: 98; Ingold 1994b: 346;

Smith 1994: 706–707).⁴ This is also the theoretical division I will draw upon. As one of the contradictions of primordialism, in the instrumentalist approach it is held that ethnic or national identity is not a fundamental, natural or inevitable outcome of ‘ancient’ ethnic groups, but rather the instrument of leaders or others vying for power, benefit and prestige (Brass 1985: 30–31).

Among others Brubaker (1996: 15) and Eriksen (1993b: 4) have argued that primordialism and instrumentalism are outdated concepts, proposing more elaborate schemes for classifying ethnicity theories. However, on other occasions Eriksen (1993a: 54) has however also written about instrumental- ism and primordialism as important perspectives on ethnicity. In any case, the terms continue to have wide currency and have been in active use throughout the 1990s (e.g. White 1999: 789; Schmidtke 1998: 25).

3 The affinity of Shils and Geertz with current primordialism is however often exaggerated, in particular that of Geertz (Calhoun 1997, 31; Foster 1991, 237). For example, Rex (1996, 82, 98, 150) states that Geertz provides the clearest statement of the primordialist account, contrasting ‘primordial’ aspects of ethnicity such as kinship, neighbourhood, lan- guage, religion and customs with personal attraction, tactical necessity, common interest or incurred moral obligation. Rex presents Geertz’s account of the former as simply

‘given’, ‘ipso facto’ and ‘unaccountable’, thus ignoring Geertz’s use of modifiers such as

‘attributed’ and ‘thought to be’, indicating that this is not how he himself sees these phenomena, but how he thinks they are seen by others (cf. Linz 1985, 204).

4 The dispute between Gellner and his former student Anthony Smith is often confused with instrumentalism and primordialism (Smith 1996b: 371). It is in fact usually more a case of that between modernism and perennialism (Smith 1994: 707-708; 1995: 35;

Gellner 1996: 366; 1997: 90, 13).

Jentleson relates views of the viability of preventive diplomacy to whether ethnic conflict is understood in a primordialist perspective, as

‘manifestations of fixed, inherited, deeply antagonistic historical identities’, or not. He asks rhetorically, ‘aren’t many of these conflicts just the playing out of history – of “Balkan ghosts” that still haunt the region, of precolonial African tribal hatreds, of other deeply historical animosities?’ (Jentleson 2000b: 5, 6). The same question is raised forcefully by Ginifer and Eide (1999b: 78).

Lack of attention to the role of the media in conflict prevention policies could be taken to imply an underlying primordialist perspective on ethnic conflict. At its most extreme, the application of a primordialist perspective to ethnic conflict entails bleak pessimism about doing anything to prevent viol- ence. In such a perspective conflict is seen as resulting more from the very nature of the groups involved, their cultures and histories, and the direct per- son-to-person interaction between them, rather than as growing out of a specific socio-political context in which factors such as elite manipulation and the media might play a catalytic or even causal role. In contrast, an instrumentalist approach entails taking the media seriously, because at least in principle it is possible to alter the context of conflict and counteract the manipulation that propels the conflict forward.

Anderson’s (1983) ‘imagined community’ approach to ethnic groups is relevant for the point I am making about instrumentalism here. While Anderson, unlike instrumentalist theoreticians, does not say much about the actual driving forces behind ethnic identities, his popular explanation of how they function in modern, large-scale societies combines well with an instru- mentalist perspective.

Anderson’s main point concerns the way in which standardised culture facilitated by new technology and social organisation becomes a medium through which feelings of community can be extended from the sphere of inter-personal contact to larger and more abstract groups.⁵ Because he is looking at the appearance of nationalism as a modern phenomenon over the recent centuries, his focus is on the role of printing, standardised languages and universal education. In our times, however, the mass media are perhaps the strongest pillars of imagined communities – also when they go to war with each other:

Through the mass media, all those belonging to an ethnic group, irrespective of geographic location, can become involved in political events. In situations of eth- nic tension and conflict the mass media can thus, in principle, convey at speed a succession of hostile images, irrespective of distance. The production and con- stant renewal of these images serve to reinforce the identity of one’s own group and exclude the community seen as the enemy (Reljic 1999: 18).

5 Anderson’s argument in turn feeds heavily off Deutsch’s (1966: 86-152) work on social communication, mobility and culture – without acknowledging so (Coakley 1992: 10;

Kellas 1991: 60). It also bears a striking resemblance to both Gellner’s preceding (1964:

69) and contemporaneous (1983) work (Smith 1992: 3).

The OSCE originated in the Helsinki Process, a series of meetings incepted in 1972, and then officially entitled the CSCE (Conference on Security and Co-operation in Europe) in 1973 (Glover 1997: 166; Berthelsen 1995: 3).

This precarious process of negotiations between East and West was facilita- ted by several factors. As its location, Finland provided perhaps the most truly neutral of any middle ground between the two sides. Paradoxically it was perceived by the West Bloc as a process to overcome the borders divid- ing Europe, while the East Bloc saw the mere initiation of a joint process as an implicit confirmation and acceptance of those borders, in particular of those dividing Germany (Kemp 1996: 5–6; Glover 1997: 166; Höynck 1995:

15).

The CSCE entered a metamorphosis at the first opportunity after the watershed events of 1989. At the 1990 Paris Summit a new course was staked out in the Charter of Paris for a New Europe, and the CSCE rapidly became a larger and more important institution (Höynck 1995: 22). How- ever, it was only decided at the 1994 Budapest Summit to update its name to the ‘OSCE’ to reflect those changes. The old CSCE had been constructive, but largely hemmed in by the divide between the two Cold War blocs. Espe- cially after the adoption of the so-called Helsinki Final Act in 1975, it had focused on human rights as a part of what it called the ‘human dimension’.

In contrast, the new OSCE took up a position closer to the centre of Euro- pean politics.

There were several reasons for this newfound potency in the 1990s.

Firstly, the OSCE could draw on the image of the Helsinki Process in both East and West as a relatively neutral common ground, not tarnished by the domination of either of the previous blocs (cf. Kemp 1996: 1). Thus Russia tended to favour the OSCE as a less threatening alternative to NATO (The Economist 1999: 36; Ivanov 2000: 14; Zhurkin 2000: 65).⁶ Correspondingly, Western states tended to favour the OSCE among other reasons because it has been the least obtrusive vehicle for entering the East European sphere.

Secondly, unlike organisations such as the EU and the Council of Europe, the inclusion of both the USA, Canada and Russia in the OSCE gave it a firmer base in global Realpolitik for dealing with the security issues of post- Cold War Europe (Höynck 1995: 13; cf. Väyrynen 1996: 29).

Thirdly and finally, historical contingency in the form of the dissolution of Yugoslavia played an important role. Because Yugoslavia had never entirely succumbed to Soviet domination, upon its collapse it became a rela- tively neutral middle (battle)ground between Russia and the West, where heavy Western field operations were far more feasible than in Chechnya, Nagorno-Karabakh or even the rapidly Westernising Baltic countries. It is

6 |The OSCE’s image as neutral should however not be exaggerated. In practice it has con- sistently performed a benevolent civilising mission on behalf of the Western states. This is illustrated by the fact that election observation missions have only been sent to former communist countries, to examine, in the educational sense of the word, democratic pro- gress.

not incidental that the large-scale OSCE peacekeeping operations envisaged by among others Vetschera (1997: 155) in Nagorno-Karabakh have never materialised. Thus the OSCE grew rapidly through the escalating sequence of impromptu post-Yugoslav operations (however belated and failed, cf.

Väyrynen 1996: 28). Had events in Yugoslavia not taken the turn that they did, the OSCE would certainly have been an entirely different organisation today.

In fulfilling its role as an increasingly central multilateral actor in Euro- pean security politics in the 1990s, the OSCE had conflict prevention prede- fined as one of its main tasks through the CSCE Helsinki Process. Most importantly, the Charter of Paris from 1990 mentions conflict prevention explicitly (p. 6), and Chapter III of the Helsinki Decisions from 1992 is wholly devoted to early warning, conflict prevention, crisis management and the peaceful settlement of disputes (cf. Lund 1996: 7; CSCE, 1992).

The OSCE has been involved in several important events combining local media and conflict prevention in the former Soviet Union. Along with seve- ral other organisations, the OSCE supported the first representative meeting of Caucasian journalists in Kobuleti in April 1996, which was followed up with new meetings in subsequent years (Navarsardian 1999: 68). In 1997 it helped organise several meetings between Georgian and Abkhaz journalists in connection with conferences sponsored by the Sakharov Foundation in Moscow and the OSCE in Warsaw (Pachkoria 1999: 56). A similar meeting between Georgian, Abkhaz and South Ossetian journalists was arranged again in Warsaw in 1999 (OSCE 1999a: 34). At that meeting the participants undertook initial steps to establish the All-Caucasian Union of Journalists. In 1999 the OSCE also organised the first regional meeting of media actors in Central Asia, which will be further discussed below.

However, in order to fully understand the place of these media-related events in the OSCE’s general conflict prevention strategy, it is necessary to place them within a broader perspective on the OSCE. Thus, in the following sections I will go on to look in greater detail at the OSCE and its work, in an attempt to uncover the progress of policy priorities and activities over time.

One way of getting a simple overview of the OSCE’s attention to media issues is by looking at the degree to which it discusses them in its official documents. Every year since 1993 the CSCE/OSCE has produced an annual report that is easily available. Each report sums up the activities of the orga- nisation’s various parts during the past 12 months, thus providing a cross- section of the development of its foci of attention over time. In order to detect changes in the organisation’s focus on media affairs, I have found that the frequency of the word ‘media’ is the most reliable indicator. It would also be possible to combine this with the frequencies of a number of words, such as ‘television’ and ‘journalist’, but in practice these tend to mirror the frequency of ‘media’ and therefore add little new information.

1993 1994 1995 1996 1997 1998 1999 2000

Times ‘media’ mentioned Number of words in document

FREQUENCY OF THE WORD ‘^MEDIA’ AND TOTAL NUMBER OF WORDS IN OSCE

ANNUAL REPORTS. FOR A FULL OVERVIEW OF THE DATA THAT MAKE UP THIS GRAPH,

SEE APPENDIX 2.

The graph above gives a simple representation of the change in mentions of

‘media’ in OSCE annual reports over time, relative to the increasing length of the annual reports. There has been a steady increase in the number of times ‘media’ is mentioned in each report, from 0 in 1993 to 176 in 2000.

At the same time, however, the number of words in each report increased from 6,916 in 1993 to 72,984 in 2000. While the increased length of the annual reports in itself should result in a greater gross number of ‘media’

mentions, it is still not enough to explain the increase that actually took place. This is because the increase to some extent is due precisely to the addition of media issues to the OSCE agenda, and has resulted in a higher proportion of the reports being taken up with media issues. In addition, the mentions of ‘media’ have increased more than the word length of the reports during the corresponding period. While the length of the reports increased by almost six times, ‘media’ mentions increased by more than 43 times.

The flattening out in mentions of ‘media’ from the 1998 Annual Report to the 1999 Report can be ascribed to the fact that the Representative on

Freedom of the Media (RFOM) got particular attention in 1998, since this was the first year in which that office functioned fully. Thus, while there were similar numbers of mentions of ‘media’ overall in 1998 and 1999 (109 and 104 cases) at the same time as the length of the annual report increased by about 25 per cent, the RFOM was associated with 72 of the mentions in the 1998 Report, and only 44 of the mentions in the 1999 Report.

The classification of ‘media’ mentions in relation to other topics on the basis of the surrounding text is such an unreliable and subjective enterprise that those numbers are not included here. In particular, attempts to determine whether ‘media’ mentions are tied directly to conflict prevention or ethnic conflict founder due to the difficulty of defining such ties and the scarcity of such mentions. Tentative numbers are nonetheless included in Appendix 2.

What the numbers and graph indicate, is that, judged by the synopsis of OSCE activities provided by the annual reports, the organisation has increased its attention to media issues. Of course, the evidence provided by such a frequency analysis can be no more than circumstantial. Among other things, the possibility must be taken into account that the various parts and activities of an organisation such as the OSCE do not necessarily receive attention and space in official organisation-wide documents such as the annual reports proportional to their importance for activities on the ground.

However, at the indicative level at which it is intended, the direction in which the word frequency analysis points is still clear: during the 1990s, the media attained greater prominence in the work of the OSCE. In the next sec- tion I will further examine and compare the HCNM and the RFOM. It was the latter that caused the great jump in ‘media’ mentions from 1997 to 1998, and around which the graph above revolves.

It is particularly interesting to compare and contrast the HCNM (High Com- missioner on National Minorities) and the RFOM (Representative on Free- dom of the Media) since they deal with the two main fields under consider- ation in this article: conflict prevention and the media. On the one hand, the HCNM is one of the main original pillars of the OSCE’s conflict prevention platform. On the other hand, the establishment of the RFOM indicates a strengthened emphasis on media-related work in the OSCE in the second half of the 1990s. Thus a joint examination of the two should make it pos- sible to say something about the relationship between the OSCE’s conflict prevention and media activities.

It was decided to establish the HCNM at the 1992 Helsinki Summit (CSCE 1992: 7). Max van der Stoel became the first HCNM, and still occu- pies the post at the time of writing. The office is located in the Hague in the Netherlands. From the very outset, preventive diplomacy was defined as one of the HCNM’s main functions, and the HCNM is in turn considered one of the OSCE’s main instruments of conflict prevention (Berthelsen 1995: 4;

The Economist 1999: 36). When the Helsinki Decisions from 1992 intro- duced conflict prevention as a central task of the evolving CSCE process, they also established the HCNM as one of the main instruments for carrying out that task (Vetschera 1997: 148; CSCE 1992: 7). Strengthening the HCNM has subsequently been seen as a way of strengthening the OSCE’s conflict prevention capability (Väyrynen 1996: 23).

The activity of the HCNM in the Baltic countries is particularly well- known. He played a central role in dampening the conflicts between the Estonian and Latvian populations and the Russian minorities. Apart from the Baltic countries, the HCNM has mainly been active in Central Europe (e.g.

Slovak–Hungarian relations, Roma affairs) and the former Yugoslavia, and less so in the CIS (Commonwealth of Independent States). His work has ten- ded to focus on possible majority infringements of the rights of minorities, and not so much on the role of the media in engendering possible conflicts between ethnic groups. The HCNM generally works by first visiting an area, discussing local problems with governmental officials and representatives of ethnic groups, and subsequently making recommendations for alterations of legislation and practices. Throughout this process he normally strives to be discrete, and does not engage in formal mediation (Peck 1999: 110; Kemp 1996: 19). The official mandate of the HCNM does not even explicitly spe- cify media-related activities on his part, mentioning them only once as one of several possible sources of information on minorities, not something to deal actively with (CSCE 1992: 11, § 23a).

The office of the RFOM was established five years after that of the HCNM. The decision to set it up was taken at the Lisbon Summit in Decem- ber 1996, and the office officially started work at the end of 1997, but was

not fully operative until 1998 (Reljic 1999: 26; OSCE 1998a: 4, 57).⁷ The post was filled by Freimut Duve, who still holds it at the time of writing. It has a staff of about six in addition to the Representative himself, and is loca- ted in Vienna, along with the majority of the OSCE institutions (OSCE 1999: 1).

While the HCNM has focused geographically on the Baltic countries and Central Europe, the RFOM has looked further afield, covering the entire for- mer Soviet area as well. In his work there, he has focused on informing media workers about their rights and protesting to governments about their abuses of those rights, and less on ethnic hate speech.

One of the main exceptions to that rule was the RFOM’s visit to Mace- donia in April 1998, where he discussed the use ‘of the media to create eth- nic tensions’ and ‘pit group against group in this politically tense area’

(OSCE 1998a: 46). However, that occurrence does not reflect most of his activities during the 1990s, and in any case Macedonia falls outside the geo- graphical scope of this article.

One of the main events organised by the RFOM so far was a conference in October 1999 in Kyrgyzstan. In organising the conference, the RFOM co- operated with the OSCE Centre in Bishkek, the Union of Journalists of the Kyrgyz Republic and the Government of Kyrgyzstan. The conference gath- ered more than 80 journalists from four of the five Central Asian countries to discuss the media in Central Asia. Unsurprisingly – considering the com- ments earlier on in this article regarding the authoritarian media policy in Turkmenistan – that was the only one of the Central Asian countries which was not represented.

The conference focused on the following topics: freedom of the press as a fundamental human right; problems of safeguarding the rights of journalists in Central Asian states; legislation and Central Asian media practices and their conformity with the principles of freedom of speech; problems of access to information; the role of independent journalist unions and associ- ations in safeguarding the rights of media representatives; principles of pro- fessional co-operation and the mass media in conflict situations (OSCE:

2000c: 239; OSCE 1999a: 44–45).

It is noticeable that out of the seven conference topics, only one was explicitly related to the role of the media in conflict. Thus the general focus of the conference was on the rights of journalists and the media, rather than on their responsibilities and ability to influence situations. Human rights and freedom of the press are of course particularly topical issues in Central Asia, requiring immediate attention. However, taking into consideration the fact that there are also serious ethnic and religious conflicts in the region, in par- ticular related to the civil war in Tajikistan, one might argue for at least greater parallel emphasis on the potential and actual roles of the media in those conflicts. A similar conference was organised in Tajikistan in Novem- ber 2000. Again there was an overwhelming focus on freedom of speech and the rights of journalists.

A perusal of the RFOM’s ‘interventions’, that is to say actions, also shows up an overwhelming emphasis on the same topics. Many of the inter-

7 In the 1997 Annual Report budget for 1998 there was still no separate posting for the RFOM, which goes to show how little advanced the plans for the office still were (OSCE 1997: 50).

ventions are letters to the authorities of various East European states, protest- ing the arrest of journalists or misapplication of libel laws to control unruly media outlets. Few or none are explicitly related to hate speech or conflict prevention (though a letter protesting the arrest of a journalist may in fact indirectly be aimed at the authorities’ attempts to silence the journalists of an opposing ethnic group) (OSCE 2000c: 275–283).

Two of the most important media-related conflict prevention events in the former Soviet Union during the 1990s were in fact not even initiated by the RFOM: a seminar on conflict in the Caucasus and the role of mass media organised by the OSCE Mission to Georgia in co-operation with the Council of Europe and the Black Sea Press Agency, and the setting up of the All- Caucasian Union of Journalists by the ODIHR and the OSCE Mission to Georgia.

There is a striking contrast between the title of the HCNM and that of the RFOM. The HCNM is emphatically the High Commissioner on Nationali- ties. It is often highlighted up front in official presentations that this means that the HCNM is not an ombudsman or spokesperson for the minority groups and their rights, but rather an arbitrator between all parties to inter- ethnic conflicts – having equal relationships with minorities and majority governments (OSCE 2000b: 1). This has been particularly important in con- nection with the HCNM’s activities in the Baltic countries (cf. The Econom- ist 1999: 36).

In contrast, the interest of the media is defined into the title of the RFOM:

Representative on the Freedom of the Media. Though the RFOM similarly to the HCNM is a representative on, the title of the office is clearly slanted towards the rights of the media, rather than an independent position between the media and the authorities, between the rights and duties of the media, and between the media of antagonistic groups. In addition, the mandate of the RFOM is saturated with human rights and press freedom discourse, and con- flict prevention is not mentioned once (OSCE 1998b). Early warning is, however, mentioned in the mandate of the RFOM, but explicitly in relation to ‘freedom of expression and free media’ (OSCE 1998a: 121).

Thus it appears that the RFOM has first and foremost perpetuated the OSCE’s human rights heritage from the CSCE and the Helsinki Process, by focusing on the rights of the media against violence, censorship and infringe- ments of freedom of speech. However important this aspect of the RFOM’s work, it seems that it to some extent has come to overshadow work related to the role of the media in conflict, and the responsibilities of the media to soci- ety.

1993 1994 1995 1996 1997 1998 1999 2000 2001 HCNM RFOM

HCNM AND RFOM PERCENTAGES OF TOTAL OSCE BUDGET (DATA IN APPENDIX 3).

In addition it seems that the RFOM as a branch of the OSCE has had a lower financial priority than the HCNM. Firstly, as mentioned, it took five years after the HCNM was established before the RFOM was even set up. Sec- ondly, during the period from 1998 to 2001, when the two institutions have existed side by side, the RFOM has consistently received less financing than the HCNM.

One might think that this is due to the later establishment of the RFOM, and that there is simply a time lag in the development of funding for the RFOM. However, not only has the RFOM had more limited financial means available than the HCNM each year, but the budget of the RFOM has also at all times constituted a smaller percentage of the total OSCE budget than has ever that of the HCNM. That is to say, even in 2000, when the RFOM’s per- centage of the overall OSCE budget reached a high of 0.5 per cent, it was less than the minimum the HCNM has received, namely 0.65 per cent in 1998. During its first year of existence the HCNM in fact received 2.17 per cent of the budget.

Another tempting counter-argument is that due to the late establishment of the RFOM, the ODIHR was already carrying out much media work in connection with its election monitoring, and has continued to do so. Admit- tedly, the RFOM and the ODIHR have arrived at a division of labour by which the ODIHR remains responsible for dealing with the media in elec- tions and includes media monitoring units in all of its election monitoring missions, while the RFOM deals with ‘structural issues’ (OSCE 1998a: 49;

OSCE 1997: 27). However, this cannot entirely explain the smaller budgets of the RFOM. Much of the HCNM’s work in relation to national minorities is similarly complemented by other parts of the OSCE, including the ODIHR and the CPC (Conflict Prevention Centre), which have been designated an official support function in relation to the HCNM (Kemp 1996: 19).

Furthermore, the role of the ODIHR certainly cannot explain the RFOM’s possibly greater attention to freedom of the press and human rights over the role of the media in conflicts, since this is what the ODIHR inevitably foc- uses on too. This is not to say that the ODIHR’s media monitoring and coun- selling never touches on the role of the media in ethnic conflict. Occasion- ally it does, but it is definitely not its focus.

To sum up this comparison of the HCNM and the RFOM, the HCNM has had greater priority within the OSCE, both in terms of financing and timing, and the work of the HCNM has been more explicitly slanted towards conflict prevention than has that of the RFOM. The RFOM in turn has focused on the human rights aspect of the media, and freedom of speech in particular.

conflict prevention and the media

The media have played a pivotal role in ethnic conflict in the former Soviet Union, and the OSCE initially did not include them to a corresponding ex- tent in its conflict prevention strategies. There are several possible reasons for this dissociation of media action and conflict prevention in OSCE affairs.

It may in part be traced to the academic literature on conflict prevention, which does not pay much attention to the media either. A random but indica- tive example is provided by looking up the words ‘media’, ‘news’, ‘newspa- per’, ‘radio’ and ‘television’ in the five indexed books on conflict prevention and resolution that happen to be lying on my desk (Peck 1999; Miall, Ramsbotham and Woodhouse 1999; Davies and Gurr 1998; Arbatov, Chayes, Handler Chayes and Olson 1997; Jentleson 2000a). Three of the books have no entries on any of the words. In Miall, Ramsbotham and Woodhouse there are two entries on ‘media’. In Peck there are two entries on ‘media’ and one on ‘radio’ [broadcasts]. Considering the numbers of entries on other terms that is not very much for these five central media words (e.g. in Peck 1999: four of ‘technical assistance’, seven of ‘political process’, 17 of ‘legal issues’).

Both the academic literature on conflict prevention and OSCE practice may also have a tendency to focus too much on the leadership level. Lund (1996: 113) argues that when ‘governing elites express exclusionary ideolo- gies … and use rhetoric that dehumanizes the “out” group’ it may be a local antecedent of possible genocide’ (my italics). Similarly, Stedman (1995: 18, 20) writes that ‘the humanitarian tragedies of today were caused mainly by leaders who were interested neither in reaching non-violent resolutions to conflicts nor in making concessions’ and that ‘preventive diplomacy and conflict prevention do not lessen the difficulty of choices for leaders, nor do they really lessen costs’ (my italics). Thus it is not enough to mediate between leaders and the official representatives of states and minorities.

Although leaders are necessarily crucial in formulating inter-group relati- ons, they also depend on the media to do so. In the end inter-ethnic conflict is not simply about predetermined primordial ethnic groups clashing together at the behest of nationalistically minded leaders.

Conflicts aren’t just about what people do to each other. They are also about what one group of people think or feel that another group of people are doing, or trying to do, or wanting to do … In a context of suspicion, not only actions, but also presumptions about the intentions and meanings behind actions or (inacti- ons), play a central role … This interpretive element in conflict behaviour shapes subsequent action (Ross, cited in Saunders 1998: 860).

Therefore,

The world needs ways of changing conflictual relationships both to prevent con- flict and to end it, not just ways of negotiating agreements. Changing conflictual relationships requires focusing on the human dimensions of conflict, not just on the objective interests over which enemies ostensibly fight’ (Saunders 1998:

859).

Another reason for the OSCE’s limited capacity to deal with local media in the former Soviet Union may be the fact that field staff often do not speak local languages apart from Russian. Thus media outlets using Estonian, Abkhaz or Azeri may not be directly accessible to them.

Western observers/OSCE staff have probably developed a deeper under- standing of the role of local media in inciting inter-ethnic hatred in the for- mer Yugoslavia than in the former Soviet Union (e. g. Carruthers 2000: 49–

50). This can to a great extent be explained in terms of the intense contact with the former Yugoslavia through the exceptionally large multilateral ope- rations there. The sheer numbers of staff employed in those operations, as well as their longevity, has ensured that a larger number of international staff than usual have got to know local communities in depth, and have in many cases learned Serbo-Croat. That has in turn made it possible for them to arrive at a better understanding of the actual workings of public opinion and inter-group animosity.

In contrast, multilateral operations in the former Soviet Union have been more tenacious, smaller-scale and have had a greater turnover of employees.

Perhaps most importantly, the international staff in post-Soviet field opera- tions usually speak Russian and not any other of the post-Soviet languages.

For many of them English is already a second or even third language, and Russian an arduous third or forth. A German employee may for example have had French as a second language at school and her current working lan- guage English as a third language, and have learned Russian later on. If she then finds that Russian is sufficient and ideal to communicate across groups and countries in the former Soviet Union, the prospect of learning for ex- ample Armenian will most likely be daunting. Thus, however highly quail- fied otherwise, she will remain disconnected from Armenian-language news- papers and broadcasters – not to mention Azeri ones.

It should also be taken into account that the OSCE was an organisation undergoing rapid changes throughout the 1990s, only moving on from the CSCE and the Helsinki Process during the early 1990s. The picture that should be drawn is therefore perhaps not one of an organisation badly de- signed for its tasks in the conflict prevention field, but rather one of an orga- nisation in the making, through interaction with a wide and constantly changing range of possible tasks.

A more constructive approach is to look at what more the OSCE could do in relation to the media and conflict prevention. While the literature on con- flict prevention tends to overlook the importance of the media, there is a separate body of literature focusing on the media that can provide some sug- gestions for further action (e.g. Eknes and Endresen 1999; Carruthers 2000).

While this literature mostly does not pay particular attention to conflict pre- vention, some of its points are applicable to the field.

For example, Eknes and Endresen (1999: 52) identify a number of local media support activities that are relevant for conflict prevention: program- ming (both news and drama) aimed at modifying people’s perceptions of a